Wireless communications deployment in industry: a review of issues
Transcript of Wireless communications deployment in industry: a review of issues
www.elsevier.com/locate/compind
Computers in Industry 56 (2005) 29–53
Wireless communications deployment in industry:
a review of issues, options and technologies
Esteban Egea-Lopez, Alejandro Martinez-Sala, Javier Vales-Alonso,Joan Garcia-Haro*, Josemaria Malgosa-Sanahuja
Department of Information Technologies and Communications, Polytechnic University of Cartagena,
Campus Muralla del Mar s/n, E-30202, Cartagena, Spain
Received 6 December 2003; accepted 15 October 2004
Available online 13 December 2004
Abstract
Present basis of knowledge management is the efficient share of information. The challenges that modern industrial processes
have to face are multimedia information gathering and system integration, through large investments and adopting new
technologies. Driven by a notable commercial interest, wireless networks like GSM or IEEE 802.11 are now the focus of
industrial attention, because they provide numerous benefits, such as low cost, fast deployment and the ability to develop new
applications. However, wireless nets must satisfy industrial requisites: scalability, flexibility, high availability, immunity to
interference, security and many others that are crucial in hazardous and noisy environments. This paper presents a thorough
survey of all this requirements, reviews the existing wireless solutions, and explores possible matching between industry and the
current existing wireless standards.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Wireless networking technologies and industrial application; Information and communication technology in industry
1. Introduction
Enterprises are as efficient as their processes are.
Information [1] is one of the most important assets in
* Corresponding author. Tel.: +34 968 325314;
fax: +34 968 325338.
E-mail addresses: [email protected] (E. Egea-Lopez),
[email protected] (A. Martinez-Sala),
[email protected] (J. Vales-Alonso), [email protected]
(J. Garcia-Haro), [email protected] (J. Malgosa-Sanahuja).
0166-3615/$ – see front matter # 2004 Elsevier B.V. All rights reserved
doi:10.1016/j.compind.2004.10.001
the modern enterprise, because it is the basis of
knowledge management.
Current challenges are to acquire and manage more
information and to integrate and communicate all
the entities that conform the production process.
To interconnect ‘‘information islands’’, to keep large
investments and to adapt to new emerging technol-
ogies, enterprises dedicated a considerable amount
of their budgets to integrate their communication
infrastructure and information systems.
.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5330
With the arrival of Internet and new e-business
models, companies have to adopt different strategies
for doing business and be competitive. For instance,
on-line sale or Business-to-Business (B2B) [2],
require to integrate Internet operations with data
and application from backend systems. Therefore,
new technologies must be adopted, adapted and also
integrated into legacy systems. During the adoption
and integration phases, scalability and inflexibility
problems arise, in particular when there is a lack of
initial planning. In conclusion, a global view and deep
knowledge of the underlying technology are needed to
have success.
On the other hand, manufacturing and industrial
processes must provide products and services of
quality. In short, it has to be competitive. To meet this
goal, production process must have some indispen-
sable elements:
(a) F
lexibility: production must be updated con-stantly, improving products and services.
(b) Q
uality control: high levels of coordination inacquisition and analysis of data are needed.
(c) I
nventory control: new strategic trends in busi-ness, e.g., JUST-IN-TIME [3], try to decrease
fixed capital by reducing or removing products in
stock.
(d) S
peed: if a company cannot deliver on time, itmeans losing benefits, image and customer fidelity.
These elements, to a great extent, depend on the
level of interconnection among: machinery, software
(management, monitoring and control), control devi-
ces and workers. This information exchange is sup-
ported by industrial communication networks. These
networks differ from conventional ones because of
environmental constraints, mainly: resistance against
harsh environment (high temperature, corrosive sub-
stances, etc.), real-time application support, noise i-
mmunity and fault-tolerance.
As a consequence, manufacturing and industrial
processes are the main engine of development and use
of industrial communication networks.
In this continuous improvement of an industry (as
defined in ISO 9000 [4]), all the effort should not be
applied to just one goal; there are a number of tasks
that should be at the same level: mobile worker
integration into information system, data gathering by
means of instrumentation distributed in a plant,
installation of a control and logistics management
system, implementation of an Enterprise Resource
Planning (ERP) or a Customer Relationship Manage-
ment (CRM) [5,6], etc. There is a need for both
quantity and quality information flows which are
common to these improvements. Again, communica-
tion infrastructure supports these exchanges of data
flow.
As an illustrative example, fieldbusses were a
revolution in industrial communication since they
allow to interconnect several devices through a single
wire in order to control and monitor their operation.
Namely, simplifying and improving information
flows. Wireless technologies are the natural evolution
that will allow to collect new information flows (not
available at the moment) and, in addition, to improve
the existing ones.
At this point, we should question why wireless is
interesting in industry. This interest is quantifiable:
according to Forrester Research in Cambridge, 15% of
industrial companies now have wireless networks in
their plants, 6% more than the previous year [7].
Wireless technologies may provide considerable
savings in networking cost and a degree of flexibility
not known in wired systems. In addition, they can be the
solution for specific industrial problems, which are not
addressed or easily solved by traditional wired systems.
Some of the features that support wireless potential are:
� L
ower installation and maintenance costs.� S
olution for physical barrier problems inherent inwiring.
� C
omparing to fieldbusses: incompatibility betweenstandards is minimized and transmission bit rate is
increased, so that multimedia services become
feasible.
� N
ew scenarios, which add value to productionprocess, arise when communication among mobile
elements is allowed.
Of course, there are several constraints when using
wireless systems that should be carefully taken into
consideration. For instance, security issues are one of
the major concerns. Also, reliability, coverage area
and fault-tolerance are other important issues.
An insufficient knowledge of these issues may
discard wireless solutions, probably losing their
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 31
capabilities and potential. It is worth having at least a
general overview of available technologies, issues and
options.
Summarizing, productive processes depend more
and more on the telecommunication infrastructure. It
is a must to have a global overview of technology and
to have some specific selection criteria for the network
technologies: alternatives, real application cases, cost,
etc. In addition, wireless technologies are showing an
increasingly significance, due to their enormous
potential. This paper provides an overview of wireless
systems, problems, and issues regarding their adoption
in industry.
The remainder of this paper is organized as follows:
related work is discussed in Section 2. An overview of
current industrial communication networks is pro-
vided in Section 3. This section also includes an
enumeration of specific requirements for industrial
environments and scenarios of current use of wireless
in industry. Section 4 is devoted to technical details of
available and emerging wireless systems and impor-
tant general issues: benefits, cost analysis, security,
regulation, etc. Next, a description of emerging and
relevant technologies is done, providing technical
details and how each technology faces the previously
introduced issues. Section 5 deals with potential of
wireless systems in industry and shows some
application scenarios. Finally, Section 6 concludes
the article.
2. Related work
Wireless technologies interest is increasing in
industrial world. There are a number of new
organizations, consortia, conferences and workshops
that are specifically focused on industrial wireless.
Some examples are: Wireless Industrial Networking
Alliance [8], which is a consortium formed recently to
promote the use of wireless in industry; in 2003 the
Instrumentation, Systems and Automation Society
(ISA) [9] organized a Wireless Communications
Industrial Technical Conference.
A number of studies on industrial wireless can also
be found. However, most of them are related to a
specific application of wireless, such as transportation
[10], or are studies with little technical description
[11]. There are also studies regarding generic issues
of wireless applied to business (as in [12]), which
are centered on the general economic impact of
wireless.
A general overview of issues, which are relevant to
wireless adoption in industry can be found in [13], but
this is a descriptive overview, with little technical
content.
There are also complementary works, as [14],
which examines the more general relationship
between wireless and information society.
3. Communication systems in industry
The number of processes or activities that
‘‘industry’’ may encompass is probably as large as
the number of communication systems developed
within this scope. As a consequence, it is an arduous
task to provide an overview of the common features of
industrial networks. Many of the industrial commu-
nication systems have been developed to solve a very
specific problem.
A classification of industrial processes may help to
clarify the mess of terms related with industrial
communications and the nature of these systems. A
coarse division of industry would show two types of
industrial processes: management or production. A
similar classification could be done with communica-
tion systems in industry: those designed for manage-
ment and those for production.
Communication requirements and needs for the
former are very similar to those for the office
environment: Access to databases, high-data volume
exchanges and multimedia capability. Communication
networks are commonly based on Ethernet local area
networks (LANs) [15] and phone lines interconnected
by the public switching infrastructure and Internet,
i.e., the common network for the office environment.
These networks are usually organized in network
segments interconnected by switches and routers.
Large organizations may use high speed and high
performance switches.
On the other hand, production usually imposes
quite different constraints, which result in highly
specialized networks. These environments are typi-
cally high-structured and location of elements does
not frequently change. Industrial networks for
production (see Fig. 1) are arranged into hierarchical
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5332
Fig. 1. Network deployment in an industrial plant.
levels (plant, area, cell and field level), depending on
the complexity of the overall production process. Plant
level is on top, where information from lower levels
is collected and the entire automation system is
commanded by means of a Supervisory Control And
Data Acquisition system (SCADA) [16]. A plant is
divided into areas, which are made up of cell groups.
Field level is the lowest one and includes the
instrumentation: sensors and actuators.
In short, we have highly distributed architectures in
which hierarchical control modules, mainly Program-
mable Logic Controllers (PLC), are interconnected by
communication networks to provide both low-level
control functionality and data acquisition from the
instrumentation (I). Therefore, reliability and perfor-
mance of the automation system greatly depend on its
underlying telecommunication network. Note, in
Fig. 1, that the controllers could be master controllers
(MC) and slave controllers (SC).
The hierarchical network, that forms the SCADA
system and the controllers (MC and SC), is typically
made of industrial Ethernet and point-to-point
connections complying RS232 and RS485 standards.
The differences between common data networks
and industrial networks arise at the lower levels. This
is so because at cell and field level the requirements
notably change: real-time services, system availabil-
ity, and fault-tolerance become crucial. Changes are
not only functional but also physical: high electro-
magnetic interference, high temperature, corrosive
substances, etc. There is also a large number of small
devices (instrumentation) working together. The
network must guarantee requirements, such as
throughput and mean transmission delay.
However, the upper levels have little difference
with respect to the systems used for management
processes or office environments. Therefore, down the
hierarchy up to cell and field levels, the communica-
tion infrastructure is practically the same for both
management and production.
3.1. Field level
This level sets up the real difference between
common office networks and industrial networks.
The production process imposes their own specific
constraints, requirements and specifications for the
communication system at this level. The system is
closely related to the actual process under con-
sideration. As a result, there is a huge diversity in
systems.
In general, the instrumentation can be divided into
sensors (data and signal alarms on temperature,
pressure, etc.) and actuators (valves on–off, regulated
valves, speed regulators, etc.). Instrumentation has an
electronic interface to be connected to a controller.
Depending on its electronic interface, the following
coarse division can be established:
� I
nstrumentation with interface 4–20 mA.E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 33
� I
nstrumentation with interface 4–20 mA and HARTprotocol.
� I
nstrumentation with a fieldbus interface.The 4–20 mA consists of a point-to-point connec-
tion between the device and the controller through
a twisted pair cable. Information is encoded in the
amplitude of the analogue signal. The HART protocol
[17] uses modulated signal to superimpose digital
information on the conventional 4–20 mA analogue
signal. Maintained by an independent organization,
the HART Communication Foundation, the HART
protocol is an industrial standard developed to define
the communications protocol between intelligent field
devices to calibrate and control them.
Finally, the instrument can have an interface to a
commercial fieldbus system although instrumentation
with an industrial Ethernet or RS232/RS485 interface
can be also found.
In this case, the variety of interconnected devices
requires the adoption of a communication standard.
This need led to a struggle between industry consortia
to set their own proprietary fieldbusses as standard.
This situation has produced a lot of confusion
regarding current fieldbus availability and features.
At present, a wide range of fieldbusses [18] can be
found. For instance: DeviceNet, Profibus, AS-I, SDS,
Interbus, CANopen or Foundation. Since 1996, there
is also a European standard, EN50170, which is based
on three previous technologies: Profibus, WorldFIP
and P-Net. Despite this diversity, some similar
features can be found.
All mentioned busses offer transmission rates
ranging between 125 Kbps and 1 Mbps (Profibus
can reach 12 Mbps).
� D
istances between 100 and 500 m, withoutrepeaters.
� T
hey use proprietary protocols with real-timefeatures.
� L
imited message size, mostly 8 data bytes per nodeand message (again, except for Profibus).
� M
ost of them are based on RS485 interface andController Area Network (CAN) technology.
Trends in fieldbusses development are the search
for compatibility with TCP/IP and open specifications
to make easier the integration with upper levels.
3.2. Industrial environment requirements
As it was previously mentioned, the requirements
that an industrial communication system must fulfill
differ according to the level in the general structure. At
upper levels, the requirements are very similar to those
arisen in office environments. Indeed, the top level is
the boundary between office area and production area.
At this point, there will be used office solutions, whose
multimedia capabilities have a positive impact on
production management. When descending the struc-
ture, other functions (as real-time services, system
availability and fault-tolerance) become crucial. At
the field level, the most restrictive requirements
appear. In general, a communication system at this
level has to satisfy:
� I
nstrumentation with interface 4–20 mA.� R
esistance to electromagnetic interference.� R
esistance to aggressive media, such as hightemperature and corrosive substances.
� L
ow latency.� R
eal-time services support.� H
igh availability.� F
ault-tolerance.In addition, there are several problems that affect
specifically wired systems: physical barriers, difficulty
in constructing additional infrastructure associated to
wired solutions, rotary elements in machinery, pro-
ductive processes continuously moving, as well as
long distances between elements. These requirements
not only increase wiring costs but also mean inflex-
ibility when reconfiguring productive processes.
Summarizing, the requirements imposed come
from both the physical characteristics and the applica-
tions that are used within the industrial framework.
3.3. Wireless in industry
Wireless communication has been long used in
industry. The use of walkie–talkies is an obvious
example. However, its use has been limited to solve
very specific problems and rarely as the main
communication infrastructure. There are a number
of examples of wireless used in industrial activities
that can give us an idea of the role they have played
so far.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5334
Traditionally, it has been used to provide voice
services in those environments where it was the only
feasible choice, e.g., farm, forestry, mines, oil
exploitation. These systems have evolved to more
sophisticated trunking (see Section 4.7.4) systems,
which can eventually provide data transmission
services as well, becoming a more general commu-
nication infrastructure.
Point-to-point radio-links are still a common
solution in rural areas. This market is dominated by
expensive proprietary solutions.
Also satellite communications have been long used
to reach isolated and remote facilities. Mining and oil
companies have been their traditional users. More
recently, Global Positioning System (GPS) is widely
employed by a number of companies. Distribution
companies use it to track their freighter fleets.
A closer example is the utilization of Radio
Frequency Identification (RFID) [19], also called RF
Tags. These systems are used worldwide for stock
management, in stores, in airports to automate luggage
transportation, or as an anti-theft system.
In general, the aim of the system is not to support
the whole communication infrastructure but to pro-
vide or support a particular application or service.
However, emerging wireless technologies have an
increasing potential either as a solution to many
current industrial problems or even as a way of
complementing or becoming the general communica-
tion system, as next sections will try to show.
Fig. 2. Evolution from proprietary solutions.
4. Wireless technology survey
Computer science and telecommunications are
observing the exponential growth of wireless market,
which has recently experienced a notable advance.
Wireless technologies will probably become a new
paradigm in industrial development, since they are
now in commercial areas. Wireless offers low-cost
networking and the possibility of new applications.
In this section, we will try to demonstrate this
assessment by having a closer look at the character-
istics common to almost all wireless technologies.
Also, we will focus on requirements for industrial
networks, such as reliability or security. Finally, a
brief analysis of the most relevant networking
standards is offered.
4.1. General overview
Two decades ago, wireless commercial market
was scarce, and it was ruled by proprietary solutions.
Such situation is a clear analogy with current
networking industrial market (as shown in Sections
1 and 3). Fig. 2 shows a schematic view of the
evolution that made the change possible. First, new
technologies in signal processing and radio field,
plus standardized protocol design with OSI-like [20]
models shifted the situation to the development of
open solutions and standards. Then, wireless market
started to grow in a continuous symbiosis with the
needs of the users: new applications and best
features, which led to the massive wireless presence
in nowadays society.
Several standards were, or are being developed:
IEEE 802.11, DECT, GSM, etc. Each of them has its
own pros and cons. Selecting the right wireless
solution for a given problem is sometimes a complex
task that depends on many factors, namely:
� c
apacity,� r
ange,� n
oise immunity,� n
oise emission,� s
ecurity,� s
etup and running cost,� p
ower consumption,� f
ree versus licensed operation,� r
egulations,� c
ompatibility with hazardous environments,� f
ault-tolerance,� t
echnological availability,E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 35
� c
ommercial availability,� Q
uality of Service (QoS), etc.4.2. Common benefits of wireless networks
Some of their noteworthy benefits are:
(a) A
llows mobility, either to physical users or tomachinery components. For example:
� Warehouse employees can use wireless equip-
ment (such as a PDAs) to move around while
making inventory, with products statistics being
displayed in real-time in the central computer.
� In machinery, rotating parts can be commu-
nicated seamlessly.
(b) A
pplication layer software developed for wirednetworks can be directly used in wireless
environments thanks to the layered design.
(c) N
ative support for multicast and broadcastcommunications, allowing services, such as:
� Multimedia broadcast.
� Fast management information delivery, from a
master to several devices.
(d) E
asy installation, with associated savings in costand time. Such savings often compensate for addi-
tional technology expense. For example, in the
Indonesian telephonic system, a satellite launch
was preferred over dropping thousands of wires
among the 13,667 islands of the archipelago [20]. It
is not wire price, but installation cost what makes
the big difference between wired and wireless
solutions, despite significant satellite design and
launch costs.
Other situations where an easy installation is an
advantage are:
� N
eed of fast deployment, for example in cata-strophic zones.
� T
emporary networking, like business meetings.� P
laces where wired installations are forbidden, likehistorical buildings.
To quantify this benefit, we have developed a si-
mple expression for the total network installation cost
(no operation and maintenance costs are considered):
Global cost ðGCÞ¼ installation cost ðICÞ þ hardware cost ðHwCÞþ network planning cost ðNPCÞ:
GC ¼ IC þ HwC þ NPC (1)
If N denotes the total number of ‘‘stations’’, then:
IC = N � number of workers (W) � total installation
time (T) � hourly cost of one worker job (HC).
Obviously, W � T can be considered as constant. If
we name it WT, then:
IC ¼ N � WT � HC (2)
Note that WT can be calculated as ‘‘the time spent
by a single worker to install a single station’’.
Hardware costs (HwC) include the prices of all
network interfaces, wires, switches, hubs, access
points, etc. Let station cost (SC) be the average
hardware cost per station, then:
HwC ¼ N � SC (3)
Finally, network planning cost encloses those
concepts related to the design and further verification
of the network. For wireless networks, verification
phase is generally more difficult (and expensive) than
for wired ones, since coverage maps of the area to
assure good reception are needed. NPC increases as a
function of N, but not in a linear fashion (i.e., price of
planning a 40 stations network is similar to price of
planning a 41 stations network, but smaller than for a
100 stations network). Thus, if for a given number of
stations N0, the planning and verification price NPC0 is
known, we can estimate NPC function cost within an
interval of N0 using some slowly increasing expres-
sion. We heuristically selected a logarithmic formula
to characterize this function:
NPCðNÞ � NPC0 �logðNÞlogðN0Þ
(4)
Factor log(N0) is introduced to define NPC
coherently:
NPCðN0Þ ¼ NPC0 �logðN0ÞlogðN0Þ
¼ NPCðN0Þ (5)
Also, this heuristics has the desirable property of
overestimate cost for NPC if N < N0, which seems
reasonable in order to establish a comparison.
Therefore, GC is finally expressed by (6).
GCðNÞ ¼ NðWT � HC þ SCÞ þ NPC0logðNÞlogðN0Þ
(6)
For the sake of clarity, in Fig. 3 we represent the
‘‘wired/wireless cost ratio’’ versus the ‘‘number of
stations’’ for a family of input parameters, which are
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5336
Fig. 3. Comparative for the net setup cost.
selected based on average market prices. Ratio values
greater than 1 mean that wireless is a cost-effective
solution; whereas, values less than 1 mean that wired
installation is cheaper.
The following input parameters are employed:
� H
C = 40 s � W Twired = 20 min� W
Twireless = (1 h, 3 h, 5 h, and 10 h)� S
Cwired = 50 s � S Cwireless = 80 s � N 0 = 100 stations� N
PC0 wired = 1000 s � N PC0 wireless = 2000 sSumming up,� W
T is the key factor to select a wired or a wirelesssolution. Wireless is a better option as WT increases
for a wired installation: ratio grows as wired
installation process becomes slower (for example,
when building modifications are necessary) and is
less if wired installation is fast.
� R
atio increases as it does the number of stations (N),or the hourly cost rises.
� R
atio decreases as it does the wireless networkplanning cost (NC), or the station cost (SEC).
This cost computation includes only network setup.
Running cost should also be taken into account. H-
owever, it is difficult to estimate it because it depends
on quite unpredictable factors; for instance, a change
on the arrangement of machinery may create a cov-
erage fading. It is not clear whether this case offers
higher savings than a wired installation. However, t-
here are some cases that clearly offer long-term be-
nefits: when moving a facility, such a lab, previous
wired infrastructure cannot be carried; it must be r-
emoved and reinstalled with its associated costs. N-
evertheless, wireless is not affected by this issue.
4.3. Problems and disadvantages
Disappointedly, wireless systems also have several
restrictions and drawbacks, and it is necessary to know
and characterize them. Among these problems we
have:
� M
ore expensive station technology.E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 37
Table 1
� P Frequency allocation in the ISM bandsower consumption, that is specially relevant to
mobile users.
Frequency band Center frequency
� I6765–6795 kHz 6780 kHz
13,553–13,567 kHz 13,560 kHz
26,957–27,283 kHz 27,120 kHz
40.66–40.70 MHz 40.68 MHz
ncompatibility among standards that use the same
unlicensed frequency bands. For example, IEEE
802.11b uses the 2.4 GHz ISM band. Bluetooth uses
the same spectrum, and when transmitting it jams
nearby 802.11b stations [21].
433.05–433.79 MHz 433.92 MHz902–928 MHz 915 MHz
� L2400–2500 MHz 2450 MHz
5725–5875 MHz 5800 MHz
24–24.25 GHz 24.125 GHz
ess capacity than their wired equivalents. For
example, IEEE 802.11 g has a maximum transmis-
sion capacity of 54 Mbps, versus the 100 Mbps of
Fast Ethernet.
61–61.5 GHz 61.25 GHz � S ecurity issues (see Section 4.5).122–123 GHz 122.5 GHz
244–246 GHz 245 GHz
� T o guarantee coverage in any situation and locationis complicated.
� P
roblems in the co-existence of wired and wirelesstechnologies (see Section 5).
� H
idden nodes problem [22].� R
adio communication issues (see Section 4.6), suchas, noise or interference.
� H
ealth issues (see Section 4.4.3).4.4. Regulation issues
If we plan to use a particular wireless technology,
we have to contemplate three general types of
regulations, and to check if our choice verifies them.
4.4.1. Spectrum regulation issues
Equipment transmissions must be regulated to
efficiently share the finite available frequency
spectrum and avoid interference between them. Main
regulatory organizations [23] are the International
Telecommunications Union (ITU) [24], the U.S.
Federal Communications Commission (FCC) [25],
the European Telecommunications Standards Institute
(ETSI) [26] and the Japan Association of Radio
Industries and Business [27]. Additionally, interna-
tional normative may be superseded by national
regulations. Recommendations and rules apply to:
� M
aximum transmitted power, constraining therange of a particular technology.
� M
aximum permitted Electro-magnetic fields(EMF) and noise emissions.
� F
requency band.Regulatory bodies distinguish between licensed
(permission must be granted by a regulatory body)
bands and unlicensed (free) bands. A particularly
interesting unlicensed band is the ‘‘Industrial,
Scientific and Medical’’ (ISM) defined by the ITU
and used in most countries with minimum differences
(Table 1).
ISM bands are not constrained to any technology,
then any system may share the same range of
frequencies. Many wireless networks, like Bluetooth
and IEEE 802.11, operate at this spectral band since it
is free. However maximum power emission is limited
by law (in Europe, lesser than 100 mW for spread
spectrum equipment [28]).
4.4.2. Industrial and security regulation issues
Special restrictions apply to industrial areas. There
is a classification of hazardous areas depending on the
risk of accident, explosive atmospheres and environ-
mental dangerousness. Only special instrumentation,
devices and machinery with demanding security levels
are permitted (intrinsically safe equipment) [29,30].
In Europe, the Atmosphere Explosive directive
(ATEX) [31] rules the intrinsic safety for electronics.
According to this directive, wireless devices have to be
certified by the CEN/CENELEC [32,33]. USA
equivalent is FCC part 15, for radio frequency
devices, and part 18, for industrial, scientific and
medical equipment [34]. As an example, in [35] there
is a summary of regulations and standards applied to a
wide-used distributed control system in industry.
4.4.3. Radio frequency safety regulation issues
Besides, a specific technology also has to obey
severe rules about impact of radiation in human health
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5338
[36,37]. Therefore, the electromagnetic emissions
must be limited: see [38,34].
4.5. Security issues
There is a great concern about wireless implication
on security. Wireless is often seen as a threat to
communication security due to the broadcast media,
which has made eavesdropping and jamming easier.
However, we think it is a biased approach to look at
security only at this level, i.e., at data link level. It is
true that wireless transmission is easier to eavesdrop
than wired transmission, but security must be globally
treated.
A company must have a global security policy,
which includes issues such as stored data protection,
resource control access, the appropriate behavior of
the employees and, of course, data transmission
protection [39,40]. From this point of view, we think
that data transmission security should be implemented
at network layer because it provides a solution that
does not depend on the different company subnetwork
technologies. Therefore, either an Ethernet segment or
an ISDN link or a wireless link can be secured by using
the same security scheme at the network layer.
Transmission channel can be secured (at network and
upper levels) using strong security methods like IPsec
[41] (network layer), SSL [42,43] (transport layer),
Kerberos [44] (application layer) and others.
It may be argued that in industrial communication
there are a number of transmissions that do not reach
network layer. For instance, data sent from a sensor to
a PLC (which do not go through network layer) might
be eavesdropped if a wireless link is used. Those cases
are also covered since most wireless standards use
lightweight protocols (like IEEE 802.11 WEP: Wired
Equivalent Privacy) to make ‘‘wireless transmission
channel as secure as wired’’. It has been proved that
this weak protection can be broken [45] with some
effort. But, cannot wired transmission also be broken
with some effort if data are not further protected by
cryptographic means at this level?
4.6. Radio emissions issues
4.6.1. Noise and media effects on communications
Problems like noise, interference, multipath delay
and fading effects [46] are specially important in
industrial environments, because there are lots of
impulsive noise sources, lots of interfering devices,
and usual machinery distribution is prone to multipath
effects.
Multipath delay is caused by signal reflections in
obstacles. Thus, any station receives the original
signal plus several echoes of the signal. These echoes
may cancel the signal and become mere noise at
reception. Thus, the more multipath delay we have the
less Signal-to-Noise Ratio (SNR) we get. This
phenomenon is especially dangerous for frequencies
above 1 GHz (l = 1/3 m), where fast fading appears.
Fast fading produces quick and large changes in SNR
ratio if the receptor moves a distance comparable to
signal wavelength.
Transmission techniques try to avoid these handi-
caps by means of:
� E
rror detection with CRC, and error recuperationwith FEC techniques [47].
� A
daptive equalizers and other pre-demodulationsignal processing methods [48].
� A
ntenna diversity [49].� B
lock interleaving to keep away from burst errorsand other pre-modulation processes [48].
� R
obust digital modulations, such as the OrthogonalFrequency Division Multiplexing (OFDM) [48]
used by IEEE 802.11 and HIPERLAN.
� R
obust digital multiple access procedures such asspreading spectrum, or hopping sequence [48]. For
example, IEEE 802.11 uses Barker spreading codes,
a special code that avoids fading [28].
� F
ast fading can be minimized using a continuousfeedback between stations. For example, a UMTS
mobile (UE) sends the calculated SNR ratio to base
station (B-node) each few milliseconds. Then B-
node calculates the effect of fading and commands
UE to keep, reduce or increase its power transmis-
sion while B-node proceeds accordingly.
Despite all these corrective measures, channel Bit
Error Rate (BER) is usually two orders of magnitude
higher than it is in a wired system.
4.6.2. Environmental impact
On the other hand, environmental impact produced
by network interfaces could be an important issue. For
example, GSM mobiles are forbidden in hospitals and
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 39
airplanes because they can disturb electronic equip-
ments.
To some extent, this effect cannot be completely
avoided. To minimize it, manufacturers must follow
regulations about maximum permitted EMF and noise
emissions (see Section 4.4). Additionally, machinery
and equipment must be properly isolated.
Another alternative to reduce interference from
wireless network to other devices is limiting power
transmission in stations. But, this approach has two
problems: first, SNR ratio in receptors will be reduced,
which may lead to smaller data rates as well as to the
fact that real-time service may become useless.
Second, network range will be reduced. Thus, far
away devices may not reach other stations.
Additionally, firms are encouraged to execute net-
work setup by using qualified components and installers
services that are in charge to carrying out all type of
environmental impact tests to assure compatibility.
4.6.3. Health issues
In the last years, forceful discussions about health
issues have started. The aim of the discussions is to
find an answer to base station transmitter/antennas for
cellular phones, mobile phones, and other types of
portable transceivers whether are or are not a risk to
human health. It is still an open question. Several
Fig. 4. Commercial s
studies are in progress (see [37]), like the United
Nations International EMF Project [50]. Its mission is
‘‘pooling resources and knowledge concerning health
effects of exposure to EMF’’.
4.7. Networks Taxonomy and Technological
description
First, some criteria must be chosen to classify
networks. In literature, several interdependent para-
meters are used. Some of them were described in the
beginning of this section: cost, range, capacity, etc.
(See [22]).
Non-broadcasting networks can be also divided
according to their ‘‘historical evolution in market
areas and application scope’’, resulting in the follow-
ing groups:
� C
erv
ellular telephony systems.
� L
ocal loop substitutes.� T
runking systems.� I
ndoor wireless communications.� W
ireless LANS.� W
ireless PANS.Fig. 4 shows the timeline progress of the most
important systems explained in the next sections. The
ice timeline.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5340
beginning of the commercial service is shown, incl-
uding a brief explanation of the scope of each tech-
nology. Future systems, which are indented to be in
service in successive years, are also illustrated: EDGE
and UMTS.
4.7.1. Historical preview
Modern-age wireless communications begun in the
mid-19th century with Maxwell’s electromagnetic
theory, followed by Hertz’s 1880s experiments that
demonstrated EMF existence. It was not until 1895
that first commercial services appeared with Marco-
ni’s wireless telegraphy (an example of efficient low-
cost large-range system versus wired telegraphy or
telephony).
During the first decades of 20th century works
continued in both theoretical and practical tools (for
example, Armstrong’s FM or Fersenden’s Radio).
That led to the development of new commercial
services, like voice broadcasting and the like. Finally,
in the 1980s (just 100 years after Hertz’s experiments)
the first cellular systems were developed.
4.7.2. Cellular telephony systems
After an early (and unsuccessful) attempt in Japan
to create a cellular network, the first commercial
systems were put into service in the west coast of USA
in 1983. This system used analog technology, that is,
voice and signaling are transmitted over an analog
bearer. In this early system, whenever a user crossed
the cell boundary with an ongoing call, it was lost
since it did not implement communications hand-off
yet [51].
Cellular systems use space multiplexing (Space
Division Multiple Access, SDMA), i.e., different
portions of space (‘‘cells’’) use different spectral
bands, and the same spectrum bands are re-used in
different (non-neighbor) cells. To avoid interferences,
reuse of frequencies must be done as far as possible.
First generation of analog cellular systems pro-
liferated during the following years, but a set of
problems appeared:
� I
t was not easy to offer new services (such as datatransmission) due to the lack of signaling possibi-
lities.
� M
any different non-interoperable standards were inuse, like Advanced Mobile Phone Service (AMPS),
Total Access Communications Systems (TACS),
Nordic Telephone System (NMT), etc.
� S
pectrum bands were saturated, so it was necessaryto develop new modulation techniques, which use
spectrum better, and the allocation of new
frequency bands.
Later, second generation systems (2G) appeared.
This generation is characterized by using digital co-
mmunication networks, such as GSM, or hybrid IS-95
(analog/digital) equipments. Third Generation Sys-
tems (3G), such as UMTS, are intended to start service
in the coming years.
4.7.2.1. GSM. Global System for Mobile Commu-
nications (GSM) is a worldwide mobile telephony
system. Nowadays, there are GSM networks in
operation in over 160 countries, with more than 400
million users, and more than 1000 million users are
expected for 2006.
GSM started in 1982, when the Conference
Europeenne des Administrations des Postes et des
Telecommnunications (CEPT) created the Groupe
Speciale Mobile (original meaning of GSM) work-
group with the aim of creating a set of standards for the
future pan-European cellular network.
One of the keys of GSM success is its evolutionary
design. At first, a subset with reduced attributes was
developed (GSM Phase 1). Specifications were
designed to be backwards compatible, with services
and functions that began later. Subsequently, GSM
Phase 2 and GSM Phase 2+ appeared, each one
enhancing the previous ones.
GSM specifications are available in natural
language in the ETSI Website [52], but are complex
for beginners. For an introductory text to GSM, see
[51,53].
It is extensively used all over Europe and the
Commonwealth countries and its coverage can be
considered almost global. Its associated infrastructure
is extremely expensive and it is only deployed by big
telcos.
GSM uses both Frequency-Division Multiple
Access (FDMA) and Time-Division Multiple Access
(TDMA), i.e., full GSM spectrum is multiplexed in
frequency bands. These frequency bands are further
subdivided in to time-shared channels, where users
synchronously access their time-slots. GSM allows
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 41
voice service and data transmission service at
9.6 Kbps, 4.8 Kbps and 2.4 Kbps. Billing is time-
based and fees are steadily moving closer to landline
phones.
In GSM, security is provided using the ciphering
A5 family protocols [51]. ATEX terminals are
available in different companies like PACSCOM
[54]. Low rate real-time services can also be provided
through its data circuit mode. Spectrum bands are
exclusively licensed to operators and EMF compat-
ibility has to be assured.
4.7.2.2. GPRS and EDGE. GSM is a circuit switched
technology, based on TDMA multiplexing. It works
well for voice-data services because voice-data are
transmitted synchronously. Unfortunately, such
scheme is not adequate for data transmissions:
channels are exclusively assigned to a single user,
even though it does not have data to send. This leads to
inefficient channel utilization and low data rates.
To solve this problem, General Packet Radio
System (GPRS) was developed. In the near future, the
new Enhanced Data Rates for GSM Evolution
(EDGE) system will improve data capacity even
more. For this purpose, it will use new digital
modulation techniques. GPRS is a packet switched
system that uses free GSM carriers for data transmis-
sion. It allows GPRS stations to reach up to 170 Kbps.
EDGE will support up to 384 Kbps.
Unlike GSM, GPRS users are charged depending
on the amount of data transmitted. At the moment,
prices are not competitive compared to wired
alternatives.
As a GSM enhancement, GPRS and EDGE
terminals provide the same industrial potentials, and
their packet mode opens ways to new services.
4.7.2.3. UMTS. The first steps towards UMTS and
other 3G systems are found in 1985, when the
International Telecommunications Union (ITU)
announced its initiative for a Future Public Land
Mobile Telecommunications System (FPLMTS). In
the ITU 1992 World Administrative Radio Conference
(WARC) frequencies in the 2 GHz band where
identified for 3G systems. Then, the term International
Mobile Communications 2000 (IMT-2000) was
adopted for these systems. This name refers both to
the expected date of release, and to the frequency they
use. The main ITU goal was to create a broadband
solution for worldwide coverage, with support for
multimedia services. Finally, two of the proposals that
were presented for IMT-2000 systems succeeded:
� E
TSI UMTS Terrestrial Radio Access Networks(UTRAN)
� C
DMA2000, developed by CdmaONE.In January 1998, ETSI selected Wideband Code M-
ultiple Access (WCDMA) [48] as the air interface for
UTRAN. It has two modes of operation: Frequency
Division Duplex (FDD) for paired spectrum bands,
and Time Division Duplex (TDD) for unpaired bands.
ETSI continued UTRAN standardization until it
transferred this work to the international 3rd Genera-
tion Partnership Project (3GPP) [55], in the beginning
of 1999. UTRAN specifications can be consulted on
the ETSI Website [52]. For an introductory reference,
see [49].
UMTS is an evolution of GSM architecture over
WCDMA, a technique that does not divide spectrum
bands in time-slots, but in code-slots. Each user is
assigned to an exclusive code (called ‘‘spread
sequence’’), which is used to modulate its data. Since
codes are orthogonal, receptors are able to de-
multiplex single user signals by processing signal
again with the same transmission code.
One of the claims of UMTS telcos is that it can
reach up to 2 Mbps of capacity. Actually, this is very
difficult, since a single mobile must acquire most of
‘‘spread sequences’’ in a cell, with the subsequent
limitation of the amount of mobiles in the cell
(because total number of codes is restricted).
Being realistic, we can state that the great
achievement of UMTS is the increase in the users/
bandwidth ratio and a high capacity packet data
transmissions, with support for QoS services.
At the moment, the infrastructure is being deployed
and tested around Europe, but it is not available to
consumers yet.
Summing up, UMTS will offer higher over-the-air
security for industrial environments, with provision
for QoS data services and multimedia. Therefore, real-
time applications become possible. Operation through
licensed bands will ensure EMF compatibility with
environment. It is likely that ATEX devices will also
appear in the market.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5342
4.7.2.4. Industrial applications of cellular net-
works.
� Alarm systems through short messages: This type of
application uses network short message service
(SMS). Since alarms are unusual events, service
price is affordable. And since there exists global
coverage, designed solutions are universal and
reach isolated places. They are typically used in
surveillance [56].
� T
racking systems, such as fleets monitoring andmanagement, typically using GPS (see Section
4.8.2) [57].
� P
ay systems, like Ericsson e-pay solution [58].� R
emote control systems, like irrigation systems[59].
� T
elemetry [60].4.7.3. Local loop substitutes
Local loop deregulation started in 1984 when, by a
judicial sentence, Bell was split in order to encourage
free competition. Since the late nineties companies are
free to compete for the local telephony market in
Europe, USA and other countries. However, the
enormous investment needed to deploy the network
constitutes a barrier to the newcomers.
An available solution comes from wireless systems
since they minimize initial investment. The Wireless
Local Loop (WLL) business is aimed at non-mobile
users who demand broadband Internet access.
4.7.3.1. LMDS and MMDS. Local Multipoint Dis-
tribution Service (LMDS) [61] is a broadband access
used for voice, data and video-on-demand services.
Multichannel Multipoint Distribution Service
(MMDS) [61] is considered to be a complementary
technology to Digital Subscriber Line (DSL).
IEEE proposed an LMDS standard: IEEE 802.16.
Additionally, several LMDS proprietary solutions are
being deployed, such as Alcatel DAVIC physical layer
specification [62] and Motorola and Nortel Networks
DOCSIS [63].
MMDS operates on the 2.5–2.7 GHz range, under
licensed bands, providing rates of 1–2 Mbps and
covering areas in the order of 30 km.
LMDS achieves up to 500 Mbps in the aggregated
uplink, i.e., shared between all users. It operates at
frequencies higher than 20 GHz, under licensed bands,
with a few kilometer coverage and is intended to be
used in densely populated areas requiring large
amounts of bandwidth. However, LMDS propagation
characteristics are more restrictive: the higher the
frequency, the higher the attenuation. The effects of
rain and atmospheric absorption are also noticeable
and the reflections increase affecting the communica-
tions negatively.
The advantage of MMDS over LMDS is that the
former operates at lower frequencies, improving
propagation behavior and requiring less expensive
equipment. MMDS uses OFDM, which is very robust
against multipath propagation, but also very sensitive
to phase and synchronization errors. In both cases, line
of sight between devices and base station is
mandatory.
4.7.3.2. Industrial applications of WLL. WLL is
typically used in the same industrial areas as the
cellular systems (see Section 4.7.2), but with the
requirement that stations must be fixed. If the
particular application obeys this requisite, WLL offers
high capacity and reasonable cost.
4.7.4. Trunking
Trunked systems are those where the number of
clients exceeds the amount of connections that can
take place at a time. Trunking is based on the
unlikelihood that all users want to make use of
network at the same time. Public wired telephony
systems are a recurrent example of trunking systems.
Observe that this notion comprises almost all wireless
systems. For instance, using this definition GSM has a
trunking structure. Nevertheless, in commercial areas,
‘‘trunking’’ has gradually turned its meaning into
‘‘communications for professionals’’, like policemen,
taxi drivers, etc. As such they offer low cost, security,
and common needs for these collectives.
4.7.4.1. TETRA. Terrestrial Trunked Radio (TETRA)
[64,65] is an open digital trunked radio standard
defined by the ETSI [52]. It was originally developed
for public safety services, but later it evolved to a
world standard. Its goal is to meet the needs of
professional mobile radio users. As such, it is fully
oriented toward business and industrial requirements,
focused on organizational communication and as a
management tool, rather than being individuals
communication-oriented, such as GSM and the like.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 43
As a standard, it ensures interoperability between
vendors. In addition to conventional digital mobile
systems, it offers: group communications, Direct
Mode Operation (DMO): direct communication
between two, or several TETRA terminals without
the use of a trunking network infrastructure, ‘‘always
connected’’ feature, fast call setup (<300 ms), very
high security with strong end-to-end encryption, and
authentication of networks and mobiles. Data trans-
mission is offered in two modes: IP packet enabled
and a circuit mode data for specialized applications
like video surveillance or real-time services.
TETRA–air interface is conceptually similar to
GSM. It uses 25 kHz carriers, each divided into four
TDMA slots. Voice call uses one slot, and data up to
four slots to achieve high data bit rates. Within
TETRA both voice and data can be transmitted
simultaneously in different time-slots.
There are several frequency bands assigned to
TETRA: 380–400 MHz for emergency services, 410–
430 MHz reserved to civil authorities, 450–470 MHz
planned for future use, and the 806–921 MHz band.
For industrial usage, ATEX certified equipments,
like Motorola MTP700 [66] and NIROS Titan family
[67], which are available. TETRA also offers high
availability through DMO and real-time services.
4.7.4.2. Industrial applications of TETRA. Trunking
systems, and TETRA specifically, offer cellular-like
applications (see Section 4.7.2). As an advantage over
classical cellular systems competitors, TETRA offers
low cost plus support to machine-to-machine com-
munications.
4.7.5. Indoor wireless communications
With the progressive advent of wired telephony,
companies seek to private communications alterna-
tives, so that they can reduce costs and increase
workers productivity. The same reasons cause the
introduction of private indoor wireless telephony.
Both wired and wireless systems are intensively used
and demanded by firms.
4.7.5.1. DECT. Digital European Cordless Telecom-
munications (DECT) [68,69] is the ETSI standard for
private wireless indoor, and the only IMT-2000 family
member broadly used nowadays. It achieved a great
success in the market.
DECT provides for voice and multimedia traffic,
and is able to interwork with other fixed and wireless
services, such as ISDN and GSM. It uses a method
called Dynamic Channel Selection/Dynamic Channel
Allocation (DCS/DCA) that guarantees the best radio
channels available to be used. This capability ensures
that DECT can co-exist with other DECT applications
and with other systems in the same frequency, with
high quality, robust and secure communications for
end-users. For this reason, it has also attracted industry
attention.
DECT operates in the 1880–1900 MHz frequency
band, and uses TDMA like GSM. The maximum
theoretical throughput rate of multi-bearer DECT is
552 Kbps. Single-bearer throughput is 32 kbps, plus
6.4 kbps for control and signaling.
For industry, DECT confronts reliability and
availability problems by means of the DCS/DCA
mechanism. It also suits well for real-time applications
and multimedia services. Like other professional-
oriented technologies, DECT also puts special efforts
in confidential data and overall security. ATEX
compatible terminals are available from several
manufacturers, for instance ANT Telecom [70].
4.7.5.2. Industrial application of DECT. Companies
extensively use DECT for voice calls, workers paging
(alert), for messaging among different employees and/
or managers, and for extending these services in the
outdoor surroundings where the system is installed.
4.7.6. Wireless local area networks
New trends in computing devices, such as PDA, last
generation mobile phones, and, of course, mass–
market laptop usage, have led to an increasing
necessity of interfaces that allow users’ mobility
while being backwards compatible with existing
application software.
Considerable price reductions, provision for high
data rates, and unlicensed operation have made
WLANs a popular choice.
WLANs operate either in infrastructure mode as an
extension of a wired LAN, or in ad hoc mode, where
network only consists of wireless stations.
4.7.6.1. IEEE 802.11 and HIPERLAN. The most
widely used specification for WLANs was developed
by the IEEE 802.11 workgroup. IEEE 802.11 [71]
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5344
specifies a physical and medium access layers by
using spread spectrum techniques. A set of different
physical layers exists for IEEE 802.11. They are
commonly referred by an identification letter. For
example, 802.11b, 802.11a, etc.
HIPERLAN [72] is the European alternative to
IEEE 802.11. HIPERLAN Type I provides 20 Mbps
with either Frequency Shift Keying (FSK) or Gaussian
Minimum Shift Keying (GMSK) modulations accord-
ing to the transmission rate, and HIPERLAN Type II
provides up to 54 Mbps, using OFDM. Both of them
are at the 5 GHz band.
When selecting one of them to study (and possibly
deploy on an industrial environment) IEEE 802.11 has
the advantage of a longer maturity. Development kits
in the market can be easily found (at least for IEEE
802.11b); whereas, as far as authors know, there is no
HIPERLAN equipment vendor. Another advantage of
IEEE 802.11 is that it uses spread spectrum
techniques, more suitable, a priori, for industrial
environments: modulations used by HIPERLAN type
I are more sensitive to noise and multipath delay.
OFDM modulation employed by HIPERLAN II seems
to suit better, because it considerably cuts down
problems caused by multipath. Nevertheless, it forces
to carefully design the circuitry: non-linear distortion
and phase noise must be avoided in order not to loose
the orthogonality of the carriers. But, this will
probably increase the cost of the product.
Both IEEE 802.11 and HIPERLAN technologies
specify the Data Link Layer. IEEE 802.11 uses
Logical Link Control (LLC) at this level, which has an
‘‘Unacknowledged-Mode’’ suitable for ‘‘best-effort’’
service, and two other modes: ‘‘Connection-Mode
Service’’ and the ‘‘Acknowledged Connectionless
Service’’. These other two services may be more
interesting for an industrial environment. The Con-
nection-Mode Service provides a connection-oriented
service to be used in low-processing capability
devices, whose software layers do not implement
flow control and reliability mechanisms.
The Acknowledged Connectionless Service pro-
vides a scheme to confirm information delivery
without previously establishing a logical connection.
In an industrial environment, it could be very useful
in some contexts: in a network, a server usually
connects to several client devices and needs to
ensure the reception of the communication. If a
connection-oriented service is used, the Logical
Link Control software must maintain tables contain-
ing the status of each connection, becoming
impractical because of the large number of tables
that may be required. It is also useful for alarm and
emergency control because the acknowledgement of
the signal reception is required but the urgency of
the signal dissuades from previously establishing a
logical connection.
IEEE 802.11 Medium Access Control (MAC)
defines, on the one hand, a distributed control access
mechanism called Distribution Coordination Function
(DCF), and, on the other hand, an optional centralized
control mechanism called Point Coordination Func-
tion (PCF).
DCF uses a Carrier Sense Multiple Access
Collision Avoidance (CSMA/CA) suitable for asyn-
chronous traffic. Stations listen to the medium and if it
is idle, they wait a delay time before transmitting. PCF
is built on top of DCF and uses a polling scheme to
communicate with a coordinator point that controls
other devices, according to a round-robin scheme.
To prevent PCF traffic from acquiring all the
resources, a time interval called superframe is defined.
This interval has a fixed duration and it is divided into
two periods: ‘‘contention-free period’’ and ‘‘conten-
tion period’’. The former period is used by PCF to poll
the devices and does not have a fixed duration because
of the variable size of the response frame. The
remaining time of the superframe is assigned to DCF
traffic. If the medium is busy at the beginning of the
superframe, PCF will wait until the medium is idle and
the process will restart.
This mechanism is appropriate for time sensitive
services but does not guarantee the polling instant in
any case since the beginning of PCF is not bounded.
Nevertheless, the polling mechanism of PCF is similar
to the one employed by many current fieldbusses as
Profibus and can be employed together with DCF,
which would be used to notify state change messages,
and to emulate the behavior of fieldbusses while
communication characteristics are further improved.
As far as the authors know, there are not commercially
available IEEE 802.11 stations with PCF functionality
yet.
On the other hand, HIPERLAN is used to create
networks with larger radius than the coverage radius of
a single station, using multi-hop transport algorithms.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 45
Also, HIPERLAN will support QoS (and real-time)
and scheduling of different classes of traffic. IEEE
802.11 supports QoS only through PCF, and the future
IEEE 802.11e will also support QoS and real-time by
means of DCF. QoS mechanisms allow wireless
networks to emulate fieldbus behavior and to support
new services aimed at improving productivity of the
elements of the production process.
There is also ATEX compliance 802.11 equip-
ments, such as Extronics iWAP 100 [73].
4.7.7. Wireless Personal Area Networks
Overlapping the aforementioned ones, Wireless
Personal Area Networks (WPAN) technologies pro-
vide a shorter range and are intended for substitution
of wires in simpler devices. In addition, they allow to
create either LAN or ad hoc networks. These are IEEE
802.15 [74], Bluetooth [75] and IrDA [76].
4.7.7.1. Bluetooth, IEEE 802.15 and IrDA. Blue-
tooth and IEEE 802.15 operate in the 2.4 GHz band,
with 10 m (Bluetooth class II) to 100 meters (Blue-
tooth class I) range and a shared bit rate of 1 Mbps.
IrDa utilizes infrared, so, devices need line of sight.
The last issued standards provide up to 16 Mbps: IrDa
very fast IR (VFIR) (see [77]).
Bluetooth uses frequency hopping spread spec-
trum, with a common bandwidth of 80 MHz, divided
into 79 channels. Bluetooth is also a TDMA system. In
each slot, stations re-tune to a new frequency channel,
which is selected by means of a pseudo-random
sequence. Hop rate is 1600 hops per second, i.e., time-
slot is 0.625 ms.
The basic unit of networking in Bluetooth is called
piconet. It is made up of one master and up to seven
slaves. Of course, all devices in a piconet share the
same hopping sequence, which is calculated as a
function of the master clock and identity. A device
may exist in more than one piconet at a time, and may
operate either as a master or as a slave in each one.
Therefore, the networks may overlap and form
‘‘scatternets’’.
The master/slave communication is based on a poll
scheme. Slaves can only transmit in response to a
master. So, collisions appear only between different
piconets, when both select the same frequency
channel. Nevertheless, the likelihood of collision is
small (around 10�5 if the number of piconets is small)
although probability tends to increase linearly as the
number of piconets also increases [78].
Bluetooth provides two types of links between a
master and a slave: Synchronous Connection Oriented
(SCO) and Asynchronous Connectionless (ACL).
SCO packets are never re-transmitted. In an ACL
link, re-transmissions are permitted.
These characteristics let Bluetooth work on an
industrial environment, providing soft real-time
services. However, it shows some drawbacks that
generate doubts about its potential in industry: The
number of devices supported at a time is small and
coverage range is also small.
Bluetooth specifications define a series of usage
models, as, for instance, file transfer or LAN access.
At the moment, there is no usage model for industrial
environments.
In addition to point-to-point links, IrDa supports
LAN networking, defining three operation modes:
Access Point Mode, a device managing access to a
wired LAN. Peer-to-peer mode, one or more devices
forming an ad hoc LAN. Hosted-mode, two or more
devices connecting each other and a host, which
provides access to a wired LAN.
Bluetooth faces up real-time industrial require-
ments if it is used in a single piconet, and soft real-time
if ‘‘scatternets’’ are employed. They offer easy
integration with existing serial busses, through its
communication profiles.
4.8. Complementary technologies
4.8.1. RF Tags systems
These are passive devices that in presence of an RF
Tag scanner transmit a short sequence of pre-defined
bits. It is intended as an identification method and is
broadly used for stock management, in stores or as an
anti-theft system.
RFID [19] can be included in a more general class
of systems, called Automatic Identification and Data
Capture (AIDC), which is the identification and/or
direct collection of data into a microprocessor
controlled device, such as a computer system or a
Programmable Logic Controller, without the use of a
keyboard. Approaches using barcodes or smart card
technologies may be also included here.
AIDC is very popular in many service industries,
purchasing and distribution logistics, industry, man-
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5346
Table 2
Summary of features of some wireless technologies
Technology Band (GHz) Maximum bitrate (Mbps) Physical layer MAC Approximate range
GSM, GPRS 0.9, 1.8 0.17 GMSK TDMA/TDD 30 km
TETRA 0.4 0.36 DQPSK TDMA 50 km
MMDS 2.5–2.7 27 (uplink) OFDM/QPSK DOCSIS 30 km
LMDS 27–31 500 (uplink) PSK/QAM DAMA-TDMA 4 km
IEEE 802.11b 2.4 11 (DS), 2 (FH) SS/DS-DQPSK CSMA/CA 200 m
IEEE 802.11a 5 54 OFDM CSMA/CA 200 m
HIPERLAN I 5 20 GMSK EY-NPMA Extensible
HIPERLAN II 5 54 OFDM TDMA/TDD Extensible
Bluetooth 2.4 1 SS/FH TDMA/TDD 10/100 m
IrDA Infrared 16 PPM IrLAP/IrLAN 1 m
DECT 1.8–1.9 2 GFSK TDMA/TDD 50 m
ufacturing companies and material flow systems.
Automatic Identification procedures exist to provide
information about people, animals, goods and
products in transit.
The omnipresent barcode labels that triggered a
revolution in identification systems some time ago are
inadequate in an increasing number of cases. Barcodes
may be extremely cheap, but their low storage
capacity and the fact that they cannot be repro-
grammed are reducing their possibilities. The solution
would be the storage of data in a silicon chip.
The most common form of electronic data-carrying
device in use is the smart card based upon a contact
field (telephone smart card, bank cards). However, the
mechanical contact used in the smart card is often
impractical. A contactless transfer of data between the
data-carrying device and its reader is far more flexible.
In the ideal case, the power required to operate the
electronic data-carrying device would also be trans-
ferred from the reader using wireless technology.
A basic RFID system consists of an antenna, a
transmitter/receiver stage and a transponder, also
called RF Tag. The tag reacts to an external transmitter
stimulus and sends back the information that it
electronically stores.
They operate at low frequencies (30–500 kHz).
These systems usually are passive (no battery), light,
cheap and have a virtually unlimited operational life.
However, their physical range is limited (2–5 m), their
storage capacity is low, the loaded information cannot
be modified, and their data rate is low. There are also
active RF Tags (they incorporate a battery), which are
read/written devices, with considerable storage capa-
city. They operate at high frequencies (850–950 MHz
and 2.4–2.5 GHz). Their range is wider (up to 30 m)
and their data rates are higher. But, consequently, cost
is also higher. Communication protocols supported by
them are usually proprietary and cannot be easily
modified.
Some ongoing standardization projects for these
technologies are [79,80]. As RFID may be used in a
hazardous environment, ATEX compliant devices can
be acquired [81].
4.8.2. Positioning systems
Currently, two satellite systems are in existence to
provide a global location service; these are the United
States Global Positioning Service (GPS) and the
Russian GLONASS system, both military but made
available to civil users. Besides, European Union is
working on a full civilian system: GALILEO [82],
scheduled for operational use in 2008.
All these systems offer a way to determine the
absolute position on Earth with an error of a few
meters. Thus, these systems are ideal for many
tracking applications. From them, only Galileo will
also operate in indoor environments.
5. Applications of wireless systems in industry
Once known the technical details of several
wireless solutions (Section 5), the next step is
matching them with real industrial environments.
Each single case needs an individual study to pro-
perly reach the correct solutions. Nevertheless, the
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 47
following general conclusions regarding advantages
of wireless in industry can be inferred from Section 5:
� A
dopting wireless as field communication networksolves physical barrier problems inherent in wiring,
decreases installation costs, improves flexibility
when reconfiguring systems, and speeds up the
deployment of the network.
� I
t allows to minimize incompatibility betweenexisting technologies (Section 3) by providing
standardized MAC protocols. Adopting a wireless
technology implies adopting a standard, with its
associated advantages. Thus, it is easier the
integration with upper levels of the system.
� R
eal-time services are crucial to industrial auto-mation. Most of wireless network technologies
provide some type of real-time services [83,84] (see
Section 4), offering, to some extent, a solution for
the requirement of determinism. Despite they all
suffer from a degree of uncertainty, there still exists
enough flexibility to implement soft real-time
mechanisms. Koulamas et al. [85] discuss the
achievement degree of real-time requirements of
Profibus DP by UMTS, IEEE 802.11 and HIPER-
LAN, showing that at least IEEE 802.11 and
HIPERLAN are suitable to implement a wireless
system extension.
� W
ireless network bandwidth is steadily increasing(see Table 2 and Fig. 4) so new multimedia services
become feasible and system performance can also
be improved.
� N
ew scenarios, which add value to productionprocesses, arise when communication among
mobile elements is allowed. For instance, workers
provided with multimedia terminals are able to gain
continuous access to control and information
systems from any location in a factory and gain
access to any network resource all the time, e.g.,
downloading of planes and reports from a database.
These and other exposed reasons show that wireless
has a considerable potential as industrial communica-
tion infrastructure.
Of course, deploying these technologies implies
solving new problems. One of the basic difficulties to
face is the way of co-existence of wired and wireless
technologies. On this topic, a number of approaches
have been proposed [86,87], such as the interconnec-
tion of all devices to an isolated wireless network, or
use of repeaters, bridges or gateways, each of them
with their particularities. It must be ensured that, in
spite of delays generated by interconnection nodes,
temporal requirements are satisfied and performance
is maintained. It must also be taken into account the
capability of adding wireless interfaces to devices.
Despite these and other issues regarding migration
and co-existence of technologies, the fundamental
problems related to the deployment of wireless in
industry are interference and multipath propagation.
Some research carried out [88,89] is promising
regarding the performance achieved in these environ-
ments.
The area extension in which the system will be
deployed also determines the selection. Cellular
communication and systems, such as GSM, GPRS
or UMTS and trunking, may be appropriated when the
area is not bounded since, despite higher operation
cost (time or data billing), there is no installation cost.
However, WLANs, WPANs and indoor wireless suit
better for bounded areas where installation may be
assumed by a company.
In the search for integration of wireless technol-
ogies with classic industrial networks it must be also
mentioned the R-Fieldbus [90] European initiative. It
contributes to the development of an architecture for a
wireless fieldbus which would integrate current
industrial networks with emerging wireless technol-
ogies. Other approaches have been [91] and the
standard ANSI/EIA 709.1 [92].
Providing a global rule to select a standard in any
case makes no sense. Instead, some specific
approaches to the problem of matching technologies
and environments are shown in next section.
5.1. Application scenarios
Industrial activities may be classified in two very
general business processes: Management and produc-
tion. Design, engineering, purchasing, production
management and supply chain management may be
included within management. Manufacturing, physi-
cal distribution and plant and field work in general
may be included within production. Management
processes exist in all type of companies, i.e., a
manufacturing company will have a management
level on top of the other business levels.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5348
Fig. 5. Typical workflows.
Fig. 5 shows a general schema of an industry that
has to address the implementation of an ERP, a CRM,
to improve its B2B, etc. Every ‘‘box’’ generates output
data and receives input data from the other ‘‘boxes’’,
i.e., there is a relationship between entities in terms of
input/output of data streams.
Data streams must be obtained through a tele-
communications infrastructure, evaluating the possi-
bilities and benefits of using wireless as the underlying
networking infrastructure. At this point, it should be
clear that the selection of a particular wireless
technology is not a straightforward task.
5.1.1. Examples of management processes
Management processes get a crucial support from
ERP, CRM systems and other office-oriented applica-
tions. Therefore, wireless is to be used in an office
environment. Which is precisely the environment
these technologies have been mainly designed for and
widely tested. Their performance has proven to be
very satisfactory.
WLAN (IEEE 802.11) is here the usual choice as
the wireless communication infrastructure. Its data
rate (up to 54 Mbps) supports the high-volume data
exchange requirement for these business management
applications, showing similar performance to that of a
wired Ethernet network. When the infrastructure is
started from scratch a considerable cutting down of
installation cost is achieved. But, probably, the main
advantage is the flexibility provided when an existing
network has to be extended. Running cables is time-
consuming, expensive and may require construction.
Using wireless results in a rapid deployment of the
new network segments. Adding a new network
segment is just done by connecting an access point
to the wired network, and adding a new user is only a
matter of authorization. Finally, network can be
brought to previously unreachable areas, as leisure
areas, or specific areas, according to necessities.
Bluetooth may be used for substitution of wires in
office devices or as a means of interchanging data
among different devices (a PDA and a computer).
Although these uses may be seen as minor or even
frivolous applications, they can result in a great
improvement of comfort and working conditions, and
this should conceivably lead to an improvement of
productivity.
DECT has long been used in offices and its
advantages are clear. But it may also be used together
with a Voice over IP (VoIP) [93,94] application for the
enterprise voice communications which provides two
benefits: first, it may be the beginning of the
integration of enterprise systems (trends in commu-
nication are the integration of the heterogeneous
services on IP networks); second, saving costs [95]
(companies claim up to 90% of telephony costs,
especially for long-distance calls). IEEE 802.11 can
also be used as the infrastructure for mobile VoIP
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 49
communication. Indeed, it is expected to be the largest
market for VoIP by 2006 [96].
All these technologies work in unlicensed bands.
Their common features are: flexibility, low-cost and
high reliability in this environment.
As an example of the industrial trends, there is an
increasing number of software products for ERP and
CRM from leading providers as Oracle, SAP or Siebel,
that take into account wireless technologies as well as
its requirements and constraints and how to interface
to them [97–102]. The main technologies are IEEE
802.11, Bluetooth and GSM/GPRS.
Reference [103] presents the integration of an
existing SAP ERP and a radio frequency system for a
logistics enterprise to develop a fully automated
end-to-end supply chain at its warehousing facilities.
Their benefits are: improved stock accuracy, enhanced
scheduling, reduced manual picking and increased
throughput. Another example is a wireless study
from IBM [104] where it is estimated up to 26%
productivity savings.
5.1.2. Examples of production processes
There are a lot of promising examples that illustrate
the penetration of radio frequency equipment in the
production process.
Reference [105] demonstrates that there are
commercial products and systems available for a
SCADA with radio-frequency. Also, reference [106]
shows the use of a PDA, connected to a PLC using
IEEE 802.11b, for monitoring and controlling a
process.
In references [107,108], we find a variety of
wireless instrumentation and equipment, together with
industrial requirements, that comply the appropriate
standards to operate in an industrial plant.
Tetra has been successfully used in the Hamburg
Harbor [109], where it offers voice and data service to
1000 users. It will also be used to support an
Automatic Vehicle Location service for tracking and
locating containers. But the paradigm in this case is
probably the Schneider’s Global Scheduling System
(GSS), which is based on Qualcomm’s OmniTracks
System [110]. Even though it involved an investment
of US$ 30 million, its success led to the adoption of
this type of systems by more than 1000 fleet trucking
companies. The system has evolved from a truck
tracking system to a comprehensive logistics and
management system, integrating all business opera-
tions: from customer service and relationship, truck
fleet maintenance, and employee comfort and pro-
ductivity. There are two different reasons for the rapid
adoption of wireless technology by the trucking
sector: technologically, it is the only feasible
technology and, economically, it provides a clear
and quick return of investment.
5.1.2.1. New application scenario: a shipyard. This
section shows an example of how systems previously
described help to introduce new application fields in
industry. As an example, the case of Izar shipyard will
be discussed. Izar is the main military and civil
shipbuilder in Spain [111] and the Polytechnic
University of Cartagena collaborates with the process
of automating its key tasks of the repair line. The
project to apply wireless technologies for the
improvement of the production process of a shipyard
is included in the activity of the IZAR-Carenas
Chair.
The working place comprises a considerably
large extension (650 m � 155 m), which is plenty of
metallic mobile elements, where it is impossible to
construct and very costly to install a wired system,
except for the outer perimeter of the facilities.
The repair line is in charge of tuning and fixing
ships. A critical issue is the unpredictable job
arrival. It generates a need for constant dialogue
among all company levels: from management to
production engineers, and from the latter to work-
men, suppliers, warehouses, etc. The main goal of
the repair line is that the operators work on demand.
Therefore, the workflow should change weekly, even
daily. It is not possible to properly foresee the actual
load of work for a given week. In addition, materials
and human resources are scarce and should not be
wasted.
One of the main problems arise on a daily basis:
there are more than 180 mobile workers. No computer
application is used to feedback on-line to the
production manager (for example finished job, extra
needs, run out of materials, etc.) and readapt and
reorganize the production process in real-time. Due to
the constrains inherent to this industrial environment,
the only way to overcome these problems is to develop
a custom tool to make every employee work in a
collaborative manner, i.e., integrating completely the
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5350
employee into the production process. This type of
tool and its corresponding devices might be developed
but relies on a networking infrastructure that does not
exist. The only way of achieving this goal is by means
of wireless technologies. Research effort is focused on
analyzing and evaluating the suitable technologies in
this hazardous and complex industrial environment,
planning the infrastructure and deploying it in the
working place. According to internal company
reports, repairing time may be reduced up to 20% if
any employee were permanently connected.
6. Conclusions
Wireless networking offers benefits and new
functions, but also has many drawbacks that make
it difficult to select between a wired or a wireless
solution. Also, selecting a specific wireless technology
is a delicate task, depends on many parameters (see
Section 4) and requires a deep knowledge of both
technology and its potential. In this paper it is shown
that there are technologies that may face industrial
requisites: high availability, QoS, real-time, inter-
ference, immunity, etc.
Wireless solves physical barrier problems inherent
in wiring, decreases installation costs, improves
flexibility when reconfiguring systems and speeds
up the deployment of the network. In addition, new
scenarios, which add value to production processes,
arise when communication among mobile elements is
allowed.
Future work includes a performance evaluation of
IEEE 802.11 standard in a shipyard and studying the
integration of wireless sensors networks both in the
shipyard and as a location system for ecological
containers. Wireless sensor network is a novel concept
to which a great research effort is devoted at the
moment.
Acknowledgments
This work has been partially financed by the
Spanish Research Council under the project EUREKA
LARLASC (EUREKA E! 2732 EULASNET) and
by the Regional Research Council of Murcia under
the R&D project SOLIDMOVIL (2I04SU044).
Four anonymous referees made relevant comments
that improved this paper. Note that references
from Websites have been checked on 2nd February
2004.
References
[1] Y. Malhotra, Measuring knowledge assets of a nation: knowl-
edge systems for development, United Nations, in: Advisory
Meeting of the Department of Economic and Social Affairs,
New York, 4–5 September 2003 http://www.kmnetwork.-
com/KnowledgeManagementMeasurementResearch.pdf.
[2] C. Bussler, B2B Integration, Springer Verlag, 2003.
[3] E.J. Hay, The Just-In-Time Breakthrough, Wiley, New York,
1988.
[4] ISA homepage, ISO 9000, Quality management principles,
http://www.iso.ch/.
[5] D.E. O’Leary, Enterprise Resource Planning Systems: Sys-
tems Life Cycle Electronic Commerce and Risk, first ed.
Cambridge University Press, 2000.
[6] K. Sheikh, Manufacturing Resource Planning (MRP II) with
Introduction to ERP SCM and CRM, first ed. McGraw-Hill
Professional, 2002.
[7] S. Koudsi, G. Bylinsky, Hot technologies, Fortune 145 (June
(13)) (2002).
[8] Wireless Industry Networking Alliance homepage, http://
www.ember.com/WINA.
[9] ISA homepage, http://www.isa.org/.
[10] J. Scott, M. Ghazi, E. Magableh, Final Report for
MBTC Project 2023: Impact of Wireless Data Systems on
the Transportation Systems of the Future, http://www.
mackblackwell.org/research/finals/arc2023/MBTC2023.
htm-_Toc18992860.
[11] S. Kumar, C. Zahn, Mobile communications: evolution and
impact on business operations, Elsevier’s Technovation 23
(June (6)) (2003) 515–520.
[12] R. Beaubruna, S. Pierre, Technological developments and
socio-economic issues of wireless mobile communications,
Telematics and Informatics 18 (May–December (2–3))
(2001) 143–158.
[13] Industrial Wireless sWorkshop, Industrial Wireless Technol-
ogy for the 21st century, San Francisco, July 2002, http://
www.ember.com/WINA/publications.html.
[14] A. Grantham, G. Tsekouras, Information society: wireless
ICTs’ transformative potential, Elsevier’s Futures: The
Journal of Forecasting Planning and Policy 36 (April (3))
(2004).
[15] W. Stallings, Local and Metropolitan Area Networks, Pre-
ntice Hall, 2000.
[16] CERN Website, SCADA Explanation, in: http://ref.cern.ch/
CERN/CNL/2000/003/scada/.
[17] The HART communication protocol specification, in: http://
www.hartcomm.org/technical/doclist.html.
[18] Fieldbus and Industrial Network Protocol Summary List, in:
http://ourworld.cs.com/rahulsebos/Links.htm.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 51
[19] K. Finkenzeller, RFID Handbook: Radio-Frequency Identi-
fication Fundamentals and Applications, Wiley, 2000.
[20] A.S. Tanenbaum, Computer Networks, fourth ed. Prentice
Hall, 2002.
[21] J. Zyren, Reliability of IEEE 802.11 Hi Rate DSSS WLANs
in a High Density Bluetooth Environment, Intersil Corpora-
tion Press, 1999.
[22] W. Stallings, Wireless Communications and Networks, Pre-
ntice Hall, 2002.
[23] Existing regulations to radio frequency and spectrum
allocation, in: http://www.ieee802.org/Regulatory/Existin-
g_Regulations.
[24] International Telecommunication Union Website, in: http://
www.itu.int/.
[25] Federal Communications Commission Website: in: http://
www.fcc.gov/.
[26] European Telecommunications Standards Institute Website:
in: http://www.etsit.org/.
[27] Japan Association of Radio Industries and Business Website:
in: http://www.arib.or.jp/.
[28] J. Geier, Wireless LANs. Implementing High Performance
IEEE 802. 11 Networks, Sams Publishing, 2002.
[29] Georgin Website, supplier of intrinsic safety equipment: in:
http://www.intrinsic-safety.info/index.html.
[30] Industrial Automation, Process Control and Instrumentation
Professionals Website; Applying Intrinsic Safety, in: http://
www.automationtechies.com/pdf_articles/Applying_Intrin-
sic_Safety.pdf.
[31] Atmosphere Explosible directive Website, in: http://euro-
pa.eu.int/comm/enterprise/atex/.
[32] European Committee for Standardization Website, in: http://
www.cenorm.be/.
[33] European Committee for Electrotechnical Standardization
Website, in: http://www.cenelec.org/.
[34] FCC Rules and Regulations Website, in: http://www.fcc.gov/
oet/info/rules/.
[35] The Use of Industry, Standards and State-of-the-art Technol-
ogies in the DeltaVTM Automation System, in: http://
www.easydeltav.com/pd/WP_industrystandards.pdf.
[36] Guidelines, for limiting exposure to time-varying electric,
magnetic and electromagnetic fields (up to 300 GHz), Inter-
national Commission on Non-Ionising Radiation Protection,
in: http://www.icnirp.de/documents/emfgdl.pdf.
[37] J. Moulder, Cellular Phone Antennas (Mobile Phone Base
Stations) and Human Health Website, http://www.mcw.edu/
gcrc/cop/cell-phone-health-FAQ/toc.html.
[38] Radio Frequency Safety from the FCC Website, in: http://
www.fcc.gov/oet/rfsafety/.
[39] V. Segura, et al. La gestion de la seguridad: pieza clave en la
gestion del negocio, Comunicaciones de Telefonica I + D 26
(July 2002).
[40] D. Treek, An integral framework for information systems
security management, Elsevier’s Computers and Security 22
(May (4)) (2003) 337–360.
[41] IETF IP Security Working roup, in: http://www.ietf.org/
html.charters/ipsec-charter.html.
[42] SSLv3 Specification, in: http://wp.netscape.com/eng/ssl3/
draft302.txt.
[43] Transport Layer Security (TLS) Extensions. RFC 3546, in:
http://www.ietf.org/rfc/rfc3546.txt.
[44] Kerberos: the network authentication protocol, in: http://
web.mit.edu/kerberos/www/.
[45] S. Fluhrer, I. Mantin,, A. Shamir, Weaknesses in the key
scheduling algorithm of RC4, Eight Annual Workshop on
Selected Areas in Cryptography, August 2001.
[46] J. Griffiths, Radio Wave Propagation and Antennas. An
Introduction, Prentice Hall, 1987.
[47] S. Lin, D.J. Costello, Error Control Coding: Fundamentals
and Applications, Prentice Hall, 1982.
[48] E.A. Lee, D.G. Messerschmitt, Digital Communication,
second ed. Kluwer Academic Publishers, 1993.
[49] H. Holma, A. Toskala, WCDMA for UMTS, second ed.
Wiley, 2002.
[50] United Nations, International EMF Project Website, in:
http://www.who.int/peh-emf/project/en/.
[51] S. Redl, M.K. Weber, M. Oliphant, An Introduction to GSM,
The Artech House Mobile Communications, 1995.
[52] ETSI publications download area, in: http://pda.etsi.org/pda/
queryform.asp.
[53] M. Rahnema, Overview of the GSM System and Protocol
Architecture, in: IEEE Communications Magazine, April
1993 pp. 92–100.
[54] PACSCOM Website, in: http://www.gsm-modem.biz/
GSMModem.asp.
[55] 3GPP Website, in: http://www.3gpp.org/.
[56] Systevo Website, in: http://www.systevo.com/.
[57] Personal alarms ‘‘Navtrak GPS/GSM Vehicle Tracking
System’’, in: http://www.personalalarms.co.uk/gps.html.
[58] Ericsson e-pay Solution Press Note, in: http://www.erics-
son.com/press/cannes/presskit/e-ay0201.htm.
[59] Comsat GSM Control Equipment Homepage, in: http://
www.mobilecomms-technology.com/contractors/gsm/com-
sat/.
[60] Global Data StarPoint GSM Telemetry Module, in: http://
www.telenetics.com/global_data/StarPoint.pdf.
[61] W. Webb, Introduction to Wireless Local Loop: Broadband
and Narrowband Systems, Artech House, 2000.
[62] Complete DAVIC Specifications. CD-ROM. Digital Audio-
Visual Council, June 1999. Also available at, in: http://
www.davic.org/down1.htm.
[63] L. Denney, DOCSIS 1.1 Overview, IEEE 802.16 Broadband
Wireless Access Working Group, June 1999.
[64] TETRA MoU, in: http://www.tetramou.com/.
[65] J. Dunlop, D. Girma, J. Irvine, Digital Mobile Communica-
tions and the TETRA System, Wiley, 1999.
[66] Motorola MTP700, in: http://www.motorola.com/cgiss/
emea/tetra/docs/mtp700_spec_sheet.pdf.
[67] Niros Titan Family, in: http://www.niros.com/StaticPart/
NIROS4100TitanDataSheet.htm.
[68] ETSI DECT, in: http://portal.etsi.org/dect/Summary.asp.
[69] DECT Forum, in: http://www.dect.ch/.
[70] ANT Telecom, in: http://www.anttele.com/.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–5352
[71] IEEE 802.11, Wireless LAN Medium Access Control and
Physical Layer Specifications, 1999.
[72] ETS 300652 V1.2.1, Broadband Radio Access Networks
(BRAN); High Performance Radio Local Area Network
(HIPERLAN) Type I, Functional Specification, July 1998.
[73] Extronics iWAP 100 Homepage, in: http://www.extronics.-
com/Product.asp?ProductID=792.
[74] IEEE 802.15, Wireless MAC and PHY Specifications for
Wireless Personal Area Networks, March 2002.
[75] Specification of the Bluetooth System, Core, V1.0, July 1999,
in: http://www.bluetooth.com/.
[76] IR Data Association, in: http://www.irda.org/.
[77] IR Data Association Specifications, in: http://www.irda.org/
standards/specifications.asp.
[78] A. Karnik, A. Kumar, Performance Analysis of the Bluetooth
Physical Layer, IEEE International Conference on Personal
Wireless Communications, 2000, pp. 70–74.
[79] ISO/IEC 16388:1999 Automatic Identification and Data
Capture Techniques, 1999.
[80] ETSI activity on Short Range Devices, in: http://portal.et-
si.org/erm/kta/SRD/srd.asp.
[81] DI ATEX RFID Devices, in: http://www.ruggedhandheld.-
com/embedded_rfid/index.cfm.
[82] European Satellite Navigation System Galileo, in: http://
europa.eu.int/comm/dgs/energy_transport/galileo/documents/
technical_en.htm.
[83] K.C. Lee, S. Lee, Integrated network of Profibus-DP and
IEEE 802.11 wireless LAN with hard real-time requirement,
in: Proceedings of the ISIE 2001. IEEE International Sym-
posium on Industrial Electronics, vol. 3, 2001, pp. 1484–
1489.
[84] M.G. Arranz, R. Aguero, L. Munoz, P. Mahonen, Behavior
of UDP-based applications over IEEE 802.11wireless
networks, in: 12th IEEE International Symposium on Perso-
nal Indoor and Mobile Radio Communications, 2, 2001, pp.
F-72–F-77.
[85] C. Koulamas, et al. Delay Performance of RadioPhysical
Layer Technologies as Candidates for Wireless Extensions to
Industrial Networks, in: 8th IEEE Proceedings of the Inter-
national Conference on Emerging Technologies and Factory
Automation, vol. 1, 2001, pp. 133–142.
[86] Ph. Morel, A. Croisier, A Wireless Gateway for Fieldbus, in:
6th IEEE International Symposium on Personal, Indoor and
Mobile Radio Communications, vol 1, 1996, pp. 105–109.
[87] S. Cavalieri, D. Panno, A Novel Solution to Interconnect
Fieldbus Systems using IEEE Wireless LAN Technology,
Computer Standards and Interfaces 20 (1) (1998) 9–23.
[88] T. Brunsvik, S. Kjesbu, Radiowave propagation in industrial
environments, in: Proceedings of IECON, 2000, pp. 2425–
2430.
[89] H. Ye, G. Walsh, L. Bushnell, WLAN in the manufacturing
industry, in: Proceedings of the American Control Confer-
ence, 4, 2000, pp. 2363–2367.
[90] J. Haehniche, L. Rauchhaupt, Radio communication in auto-
mation systems: the R-field bus approach, in: Proceedings of
the IEEE International Workshop on Factory Communication
Systems, 2000, pp. 319–326.
[91] D.A. Roberts, ‘OLCHFA’ a distributed time-critical fieldbus,
in: IEE Colloquium on Safety Critical Distributed Systems,
1993 pp. 6/1–6/3.
[92] ANSI/EIA-709.1, Control Network Specification, 1998.
[93] J. Ellis, J. Rahman, C. Pursell, Voice, Video, and Data Net-
work Convergence: Architecture and Design, From VoIP to
Wireless, Academic Press, 2003.
[94] Net Vision Voice Solutions and VoIP Handsets, in: http://
www.symbol.com/products/wireless/nvdataphone.html.
[95] Sage Research, New Study by Sage Research Shows IP
Telephony Productivity Gains Across Entire Organization,
February, 2003, http://sageresearch.com/Feb-IPT.htm.
[96] W. Strauss, New Report Analyzes Growing DSP Chip Mar-
kets for Internet Telephony, http://www.fwdconcepts.com/
Pages/press35.htm.
[97] Datavision Wireless Products for Implementing an ERP with
SAP or Oracle, in: http://www.datavision.com/erp/.
[98] Datavision Case Study Wireless ERP, in: http://www.
deskeng.com/articles/white_more/white2/pages/datavision/
index. htm.
[99] Wireless Mobile Suite from Defywire, in: http://www.defy-
wire.com/casestudy.htm.
[100] Oracle, in: http://www.oracle.com/.
[101] Siebel, in: http://www.siebel.com/.
[102] SAP in: http://www.sap.com/.
[103] Case Study Applying Wireless in a Warehouse, in: http://
www.epicdata.com/database/EpicData_Logisics_CS10_02.
pdf.
[104] IBM Wireless Market Study, in: http://www-5.ibm.com/es/
press/notas/2002/octubre/i/wireless.pdf.
[105] Wireless Networking Equipment and Solutions for Industry,
in: http://www.scadawireless.com/applications.html.
[106] SCADA Wireless PDA, in: http://www.avidsolutionsinc.-
com/wireless.htm.
[107] Wireless Instrumentation for Process Applications, in: http://
www.elprotech.com/elpro/index.htm.
[108] Wireless instrumentation for industry, in: http://www.save-
withaccutech.com/wireless.htm.
[109] Tetra Hamburg Harbour Case Study, in: http://www.motor-
ola.com/cgiss/emea/tetra/casestudies/hamburgharbour.pdf.
[110] Qualcomm’s OmniTrack, in: http://www.qualcomm.com/
qwbs/products/omnitracs/index.html.
[111] IZAR Shipbuilder, in: http://www.izar.es/.
Esteban Egea-Lopez received the Tele-
communication Engineering degree in
Telecommunications in 2000, from the
Polytechnic University of Valencia
(UPV), Spain, and the Master Degree
in Electronics in 2001, from the Univer-
sity of Gavle, Sweden. Since 2001, he is
an assistant professor of the Department
of Information Technologies and Com-
munications at the Polytechnic University
of Cartagena. He is a PhD candidate and his research interest is
focused on ad hoc and wireless sensor networks.
E. Egea-Lopez et al. / Computers in Industry 56 (2005) 29–53 53
Alejandro Martınez-Sala received the
Electrical Science Engineering degree
(B’98, M’2000) from the Polytechnic
University of Cartagena (UPCT) in
Spain. Since 2001, he is an assistant
professor of the Department of Informa-
tion Technologies and Communications
at the Polytechnic University of Carta-
gena. He is a PhD candidate and his
research interest is focused on ad hoc
and wireless sensor networks.
Joan Garcıa-Haro received the Tele-
communication Engineering degree and
the PhD in Telecommunications in 1989
and 1995 respectively, both from the
Polytechnic University of Catalonia
(UPC), Spain. He has been an Assistant
Professor at the Department of Applied
Mathematics and Telematics (DMAT-
UPC) since 1992, and Associate Profes-
sor since 1997. In September, 1999 he
joined the Polytechnic University of Car-
tagena (UPCT), Spain, where he is Professor of the Department of
Information Technologies and Communications. He has been
involved in several National and International research projects
related to electronic and optical packet switching, B-ISDN design
and planning, next generation Internet, wireless and sensor net-
works, value-added services and performance evaluation issues. He
was a visiting research scientific at Queen’s University at Kingston,
Ontario, Canada. He is author or co-author of more than 50 papers
mainly in the fields of switching and performance evaluation. Since
1994 he is regional correspondent of the Global Communications
Newsletter (and Editor in Chief from 2002) included in the IEEE
Communications Magazine, Associate Technical Editor from Jan-
uary 2000, and Technical Editor of the same magazine from March
2001. He also holds an Honorable Mention for the IEEE Commu-
nications Society Best Tutorial paper Award (1995).
Javier Vales-Alonso received the tele-
communications engineering degree from
the University of Vigo, Spain, in 2000.
Since 2003 he is an assistant professor of
the Department of Information Technol-
ogies and Communications at the Poly-
technic University of Cartagena, where
he is pursuing his doctorate. His current
research interest includes ad hoc and
sensor networks.
Josemaria Malgosa-Sanhauja received
the Telecommunication Engineering
degree in Telecommunications in 1994
from the Polytechnic University of Cata-
lonia (UPC), Spain. In November 2000, he
received the PhD degree in Telecommu-
nication from the University of Zaragoza
(UZ), Spain. He has been an assistant
professor at the Department of Electronic
and Communications Engineering (Uni-
versity of Zaragoza) since 1995. In Sep-
tember 1999, he joined the Polytechnic University of Cartagena
(UPCT), Spain, as associated professor. He has been involved in
several National and International research projects related to switch-
ing, multicast switching technologies, traffic engineering and Multi-
media value-added services design. He is author of several papers in
the fields of switching and multicast technologies. Since 2000 he is in
charge of the development of the new Information and Communica-
tion Technologies for the Polytechnic University of Cartagena. He is
regional correspondent of the Global Communications included in the
IEEE Communications Magazine since 2002.